Generating a synthetic population of individuals in households: Sample-free vs sample-based methods

摘要

We compare a sample-free method proposed by Gargiulo et al. (2010) and asample-based method proposed by Ye et al. (2009) for generating a syntheticpopulation, organised in households, from various statistics. We generate areference population for a French region including 1310 municipalities andmeasure how both methods approximate it from a set of statistics dervied fromthis reference population. We also perform sensitivity analysis. Thesample-free method better fits the reference distributions of both individualsand households. It is also less data demanding but it requires morepre-processing. The quality of the results for the sample-based method ishighly dependent on the quality of the initial sample.